Does Rising(falling) edge time and high time influence the determine of trigger(or siganal)[PCIE6363]

[65216]

Hi!

I use a external sample clock to count edge.

But I find something is wrong:I send 400 external TTL-like(periodicity=100ms) sigal as sample clock,but before the sample clock played,the aquisition is over,which means NI is not receiving the sampling clock correctly.

So then I use a normal edge count task to count the Sample clock signal,I find the number is totally wrong(much larger than expcted input number).

Considering my signal is genarated by a DAC(24bit,100KHz) and not standard(more like triangular wave),so I'd like to know Under what conditions does NI count upward or think it is a sample clock

here is part of my 'ugly' sample clock signal.

which you can see:rising time~0.100s,width~0.3s,MAX~2.0V,low~(-0.9V-0.3V)

the total 'ugly' peak send to counter is 2500,but the edge counter counts  more than 467852 peaks

I'd like to konw why and how to solve this?

Thanks!

[Kevin_Price · Accepted solution]

Well, I'd say that your analog signal isn't sufficiently "TTL-like" in terms of both rise/fall times and voltage level.  I know there are simple interface circuits that could be used to make it TTL-compatible, but don't have enough specific familiarity to suggest a particular one.

It looks like you're generating the pulses with a Delta-Sigma DAC -- what's the device?  What's the reason for preferring the bandwidth-limited pulse signal?   

Another option could be to use the analog triggering capability of your 6363, if the AI system isn't already spoken for.  When analog triggering is configured, there's an internal TTL signal known as the "analog comparison event" that asserts high when trigger conditions are met.  You could use that internal signal as the sample clock for your edge count task.  Notably, this signal will assert and deassert repeatedly while the AI task continues to run, even if the AI task itself only reacts to the first instance.

    In LabVIEW, you'll probably need to right-click the DAQmx terminal constant (or control) specifying the 'source' input to DAQmx Timing, then select "I/O Filtering..." and check the box to "Include Advanced Signals" first.  Then the drop-down will give you access to the analog comparison event, as well as a bunch of other internal and special signals.

-Kevin P

[65216]

Thanks a lot!

I have to use DAC to genarate signal because It's a synchronization requirements and hardware limitations of another system.

Then about your suggestion of Analog Comparison Event,Do you mean:the comparison event output signal can be directly configured as sample clock of an edge counter task?(like below but to a sample clock terminal instead of <PFI> )

And the comparison event output signal can be repeating until i stop the AI Analog Comparison task?

[65216]

Hi!

I'm trying to use the AnalogTriggerEvent to genarate good sample clock.

A new question: It seems that the triggers(or the  AnalogTriggerEvent) can't be exits without a certain (ai/ao/di/do/counter) channel.

So i have to configure a task channel with trigger,then i can routed this AnalogTriggerEvent out?

and the trigger have many different types(start,arm,reference and pause),which is the best as you see?

[Kevin_Price]

---

@65216 wrote:

Then about your suggestion of Analog Comparison Event,Do you mean:the comparison event output signal can be directly configured as sample clock of an edge counter task?(like below but to a sample clock terminal instead of <PFI> )

Well... yes it can be used directly, but no, not in the way you're proposing.  As I described earlier, the direct method is to specify the A.C.E. as the 'source' input when configuring DAQmx Timing for the counter task.   The method you wrote about would be configured in the context of the AI task, thus not allowed to *directly* push the signal into the counter task as a sample clock.

And yes, the A.C.E. signal will keep repeating whenever the triggering conditions are satisfied throughout the duration of the AI task.

-Kevin P

[edit: your most recent msg came in before I finished my response above]

P.S.  You'll need a continuous sampling AI task to make the A.C.E. keep asserting repeatedly.  You could probably choose either Start Trigger or Pause Trigger for your AI task, along with appropriate parameters to mark the desired trigger point.  Then the counter task would specify the A.C.E. as the 'source' for its sample clock.

P.P.S.  Could you generate the analog pulse with your 6363 instead of the device you're using presently?  You'd get much faster rise and fall times and much sharper transitions.  The smoothed out sinusoids you show look like an artifact from using a Delta-Sigma analog output device with built-in bandwidth limits that cut out all the higher frequencies associated with a rectangular pulse.